Since Spearman and Brown first reported in 1910 ways to quantify the reliability of data sources, researchers have shown ways to quantitatively account for the relationship between independent measures (see Brennan, 2001). Today`s researchers are fortunate to have a lot of statistical software (e.g. SPSS.B, STATA) that can instantly calculate a variety of reliability measures, such as for example. B Cronbach product-moment correlation coefficient alpha or pearson. While such statistics may be beneficial to classical theory of test scores or traditional opinions about the reliability of psychological scales, these statistics are not well suited for the analysis of data from experimental subsets. In addition, the statistical software mentioned above does not contain the standard behavioral analysis algorithms necessary to calculate the reliability of individual subsales data, such as for example. B statistics for interval intervals or average duration statistics for events. Therefore, the use of Excel is beneficial for behavioral analysts, as this program is very widespread and is designed for custom formulas and custom analyses. Most relevant to this discussion is the idea that well-designed tables can reduce the computational load of data and improve the accuracy of analysis for users who are unfamilaer of statistics or computer programming. The IOA computer described here offers an accessible, accurate, and easy-to-use Excel tool for practicing behavioral analysts. Partial match in IOA intervals. To get around the described disadvantage, associated with using the IOA algorithm for the total number, the partial concordance in intervals approach (sometimes referred to as the “average number per interval” or “block by block”) breaks down the observation period into small intervals and then examines the concordance in each interval.

This increases the accuracy of the compliance measure by reducing the likelihood that the total censuses were inferred from different events of the target responses within the observation. By breaking down the example of observation in Figure 1 into small steps/intervals (15 m intervals), the partial compliance approach calculates the IOA per interval and divided by the total number of intervals. In this case, IOA would be 50% (or .5) for interval 4, 100% (or 1.0) for intervals 5 to 14 (both agreed that 0 target response appeared at each of these intervals), but 0% for intervals 1 to 3 and interval 15. Therefore, the partial concordance approach would be calculated at regular intervals by adding the IOA values (in this case 10.5) to the total number of intervals (15), resulting in a more accurate and lower percentage (70%) of the IOA than the 100% figure obtained with the total counting algorithm. 1.Note that this machine has been tested on the basis of the IOA sample data in Chapter 5 (Improving and Assessing the Quality of Behavioral Measurement); Pp. 102-124) by Cooper, Heron and Heward (2007). For all algorithms, there was a 100% correspondence between the IOA derived values that used the computer described in this article, with those of Cooper et al. Unscored-interval IOA have been reported. The IOA algorithm with an uncored interval (also called “non-interval” in the research literature) is also more rigorous than simple interval-by-interval approaches, taking into account only intervals where at least one observer records the absence of a target response. The rationale for IOA in the Uncored interval is similar to that for IOA at point intervals, except that this metric is most appropriate for high response rates (Cooper et al., 2007).

In the sample data in Figure 2, the fifth and sixth intervals are ignored for computational purposes, with both observers having received a response at these intervals. . . .